distilbert-base-cased-hate-speech
Training: The model has been trained using the script provided in the following repository https://github.com/MorenoLaQuatra/transformers-tasks-templates
This model is a fine-tuned version of distilbert-base-cased on hate speech dataset. It achieves the following results on the evaluation set:
- Loss: 0.6837
- Mae: 1.9686
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Mae |
|---|---|---|---|---|
| 0.6857 | 1.0 | 3389 | 0.6471 | 1.9725 |
| 0.3645 | 2.0 | 6778 | 0.4359 | 1.9725 |
| 0.2266 | 3.0 | 10167 | 0.3664 | 1.9725 |
| 0.1476 | 4.0 | 13556 | 0.3253 | 1.9725 |
| 0.0992 | 5.0 | 16945 | 0.3047 | 1.9725 |
| 0.0737 | 6.0 | 20334 | 0.2869 | 1.9725 |
| 0.0537 | 7.0 | 23723 | 0.2709 | 1.9725 |
| 0.0458 | 8.0 | 27112 | 0.2667 | 1.9725 |
| 0.0313 | 9.0 | 30501 | 0.2589 | 1.9725 |
| 0.027 | 10.0 | 33890 | 0.2540 | 1.9725 |
Framework versions
- Transformers 4.22.1
- Pytorch 1.11.0+cu113
- Datasets 2.0.0
- Tokenizers 0.11.6